Twenty-five years ago, isolation of vancomycin-resistant Enterococcus faecium (VREm) was reported both in the UK and in France. Since then, VREm has spread worldwide in hospitals. Hospital outbreaks appeared to be related to the evolution since the end of 1980s of a subpopulation of E. faecium highly resistant to ampicillin and fluoroquinolones (the so-called clonal complex CC17) that later acquired resistance to vancomycin. CC17 isolates are presumably better adapted than other E. faecium isolates to the constraints of the hospital environment and most contain mobile genetic elements, phage genes, genes encoding membrane proteins, regulatory genes, a putative pathogenicity island and megaplasmids. Colonization and persistence are major features of VREm. Inherent characteristics of E. faecium including a remarkable genome plasticity, in part due to acquisition of IS elements, in particular IS16, have facilitated niche adaptation of this distinct E. faecium subpopulation that is multiply resistant to antibiotics. Quinupristin/dalfopristin and linezolid are licensed for the treatment of VREm infections, with linezolid often used as a first-line treatment. However, the emergence of plasmid-mediated resistance to linezolid by production of a Cfr methyltransferase in Enterococcus faecalis is worrying. Daptomycin has not been extensively evaluated for the treatment of VREm infections and resistant mutants have been selected under daptomycin therapy. Although control of VRE is challenging, a laissez-faire policy would result in an increased number of infections and would create an irreversible situation. Although so far unsuccessful, dissemination of glycopeptide-resistant Staphylococcus aureus with van genes acquired from resistant enterococci cannot be ruled out.